The displacement of water from a protein surface upon the binding of a drug has a significant, if not dominant, contribution to the free energy of recognition, and hence plays a significant role in determining drug potency and specificity. Despite the importance of water in mediating drug- protein interactions, commonly used structure-based models do not explicitly treat water as a molecule. Instead, they indirectly hydration effects by categorizing ligand-protein contacts as either hydrophobic or hydrophilic or by modeling water as a continuum. Neither of these approaches accounts for the finite size and directed nature of water's hydrogen bonds, the physics of which is essential for describing the hydration of the diverse environment of confined protein binding sites. The adoption of simplified treatments of water in drug design applications has been made necessary by the complexity of hydration phenomena and the lack of a molecular-based framework for its structural and thermodynamic analysis. In recent years, the PI has been instrumental in developing two methodologies that utilize inhomogeneous fluid solvation theory (IST) to map out solvation structural and thermodynamic properties of water in molecular detail in protein binding sites: 1) A hydration site analysis (HSA) approach, which forms the basis for Schrodinger LLC's WaterMap and 2) A corresponding high-resolution grid- based implementation, GIST, now available in the freely distributed AmberTools. Each of these analysis tools maps out 24 independent measures of structure and thermodynamics. In this proposal we will incorporate solvation structure and thermodynamic maps into virtual screening and lead optimization methodologies to improve our ability to identify and design compounds that bind with high affinity and specificity to a targeted member of a family of proteins. We propose to optimize and apply these methods to two important drug targets: the dopamine receptor D3, a target for the treatment of drug addiction, and the ?-OR opioid receptor, an important target for pain alleviation. We have chosen these receptors because of the challenges of targeting them specifically. Off-target binding often results in either the inability to discover viable drugs (D3) or drugs which have significant undesirable side effects (? -OR). Current methodologies have been ineffective in finding specific binders for these targets. Hence they remain drug targets of significant interest in both academic and industrial settings and the natural choice for the application of the new discovery methodologies proposed here.